PDL Overview
Proximity-dependent ligation (PDL) is a novel biomolecular technique that uses two single stranded DNA (ssDNA) aptamers to specifically bind to a desired target protein (Fredriksson et al., 2002; see Diagnostic Overview). When both aptamers are bound, they are brought together in close proximity and ligated to then form a double stranded DNA (dsDNA) product using T4 DNA polymerase (see Fig. 1).
CspZ Protein
Lambert iGEM considered beginning our diagnostic workflow with direct amplification of genomic DNA from Borrelia burgdorferi, the causative agent of Lyme Disease. However, after learning that B. burgdorferi rapidly disseminates into joint and cardiac tissue, we decided to change our approach by integrating a PDL assay that targets biomarkers present at high concentrations in a small volume of blood.
During conversations with Dr. Nicole Baumgarth, a Lyme immunologist at Johns Hopkins University, we were introduced to multiple surface proteins of B. burgdorferi. We ultimately selected CspZ over the other options for LANCET because it possessed a number of desirable characteristics (see Table 1):
| OspA | OspC | CspZ |
|---|---|---|
| Moderately conserved across Borrelia strains and species (Nayak et al., 2020) | Highly variable with subtypes that have significant differences (Arnaboldi et al., 2013) | ~80% sequence identity conserved between Borrelia species (Guérin et al., 2025) |
| Limited to expression while in ticks (Camire et al., 2021) | Abundant and reliable expression during early stage infection (Camire et al., 2021) | Production increases 5.3-fold in human blood, signifying B. burgdorferi’s attempts to evade the host’s immune system (Marcinkiewicz et al., 2019) |
| Short-term antibody presence due to production pre-infection (Wagner et al., 2012) | Decrease in antibodies shows non-persistent presence (Wagner et al., 2012) | Antibody levels in late Lyme indicate persistent presence (Rogers et al., 2009) |
| Existing aptamers are reported in literature (Tabb et al., 2022) | Strain variability inhibits aptamer usage (Arnaboldi et al., 2013) | Existing aptamers are reported in literature (Guérin et al., 2025) |
CspZ stood out as the most promising protein biomarker for LANCET because it had a high degree of genetic conservation, is upregulated in mammalian hosts, and has prolonged, chronic expression. The persistent and abundant expression of CspZ across various stages of Lyme ensured that our assay remains functional across different Borrelia strains and is viable for an extended period of time.
CspZ Protein Concentration
| Parameter | Formula | Resultant Value |
|---|---|---|
| Flagella filaments per individual Borrelia spirochete | 7 flagella inserted at each cell end × 2 cell ends/cell = 14 filaments/cell (DeHart et al., 2021; Zhang et al., 2020; Zhang et al., 2019) | 14 filaments/spirochete |
| FlaB filament length in B. burgdorferi | 10 μm (10,000 nm) (Kudryashev et al., 2009; Tanner et al., 2011; Zhang et al., 2019) | |
| FlaB subunits per filament | 10,000 nm ÷ 0.473 nm/subunit (Tanner et al., 2011) x 0.952 (Motaleb et al., 2004) = 2.01 × 104 subunits/filament | 2.01 × 104 FlaB subunits/filament |
| Total FlaB subunits per spirochete | 14 filaments/cell × 2.01 × 104 FlaB subunits/filament = 2.82 × 105 FlaB subunits/cell | 2.82 × 105 FlaB subunits/spirochete |
| Blood treated FlaB:unpermeabilized CspZ ratio | 1:1 FlaB:CspZ (Marcinkiewicz et al., 2019) | |
| Total CspZ molecules per spirochete | 2.82 × 105 FlaB subunits/spirochete × 1 = 2.82 × 105 CspZ molecules/spirochete | 2.82 × 105 CspZ molecules/spirochete |
| Number of spirochetes per blood volume in mice | 7100 spirochetes/ mL (7.10 spirochetes/µL) (Liang et al., 2020) | |
| Initial Number of CspZ molecules/µL | 2.82 × 105 CspZ molecules/cell x 7.10 spirochetes/µL = 2.00 × 106 CspZ molecules/µL | 2.00 × 106 CspZ molecules/µL |
To validate the feasibility of our PDL experimentation at physiological levels, we calculated an approximate range for the number of CspZ molecules found in the bloodstream. Using values derived from Liang et al., 2020, we found that the initial concentration of CspZ protein stands at 2.00 × 106 molecules/µL at two days post-infection (pi), which we compared to the lower limit of the PDL assay of 24,000 molecules/µL as characterized by Fredriksson et al., 2002 (see Table 2).
We modeled the protein concentration over time using a degradation constant of 0.05 molecules/µL/day, which is consistent with current literature (Alkhamis et al., 2025). Using Mass Action reaction kinetics, we determined the protein concentration in blood up to 300 days pi (see Fig. 2).
Two of the most common protein detection methods, ELISA (enzyme-linked immunosorbent assay) and western blot, lack the sensitivity needed to detect CspZ. Using the molecular weight of CspZ (25 kDa), the concentration of 2.00 × 106 molecules/µL was calculated to be 8.3 × 10-5 µg/mL (Rogers et al., 2009). The detection limit of ELISA (9.0 × 10-2 µg/mL) is well above the CspZ concentration, and western blot (~500 ng total protein) would require over 6 liters of blood to detect the protein, making both techniques unsuitable (Fuchs, 2023; Quinn et al., 2002).
Assumptions
To estimate a range for the abundance of CspZ molecules occurring in blood, we introduced several assumptions that simplified the calculation:
- Constant parameters: We assumed there was little variability in flagella count, filament length, and genetic expression between spirochetes and used the lower-bound values to underestimate the protein concentration
- Blood-treatment reflects physiological conditions: Exposure to blood serum was assumed to simulate conditions during infection, although immune responses could cause shifts in expression
- FlaB:CspZ ratio: CspZ was found to be expressed at a greater rate than FlaB, but because values were not quantified, we assumed a 1:1 ratio to avoid overestimation (Marcinkiewicz et al., 2019)
- Murine models simulate human physiology: Data from mice infected with B. burgdorferi was used to estimate human blood conditions after infection, as mice have high genetic and physiological similarity to humans (The Jackson Laboratory, n.d.)
Due to the uncertainty of our calculation from the assumptions above, the number of CspZ molecules in patient blood samples may be even higher than the estimated values.
Aptamers
Aptamers are short oligonucleotide ssDNA sequences designed to target specific molecules with high selectivity. Aptamers are similar in function to antibodies, but are more advantageous due to non-cytotoxicity, low immunogenicity, greater thermal stability, and better cost-effectivity (Prakash & Rajamanickam, 2015).
From recent research by Guérin et al., 2025 we discovered a set of aptamers with binding affinity to the different locations on the surface of the CspZ protein. Based on computational predictions, we chose the aptamers Apta11 (BBa_25XXUI8B) and Apta4 (BBa_2543XHTV) which exhibit the most stable binding to CspZ. For LANCET, we utilized pre-existing sequences, as developing new aptamers is a complex process which is not feasible for us at the high school level.
Ligation
The main aspects of the PDL reaction are protein-DNA binding and ligation. Protein-DNA binding occurs when the two ssDNA aptamers bind to distinct epitopes on the target CspZ protein, bringing them in close proximity. A short bridge oligonucleotide then hybridizes to complementary segments on each aptamer construct, aligning the DNA ends. This spatial arrangement enables ligation, in which T4 DNA ligase joins the two DNA strands to form a continuous product.
PDL Design
Aptamer Design
When developing the PDL assay, we decided to utilize pre-existing aptamers to CspZ that had undergone selection for optimized binding affinity for the protein (see Aptamers). We then designed specific linear ssDNA constructs for both aptamers (BBa_25KF6SDY and BBa_25PTVQ46), each containing 5′ and 3′ constant regions (BBa_252ZQF8B and BBa_25ESUVRE) surrounding the aptamer sequence with a 40 base pair linker region as recommended by Fredriksson et al., 2002 (see Fig. 3).
After receiving guidance from Dr. Mark Styczynski, an expert in metabolic dynamics at the Georgia Institute of Technology, we chose our linker regions (BBa_25SC2X09 and BBa_2537R5AB) to minimize the possibility of secondary structures. To allow the bridge sequence to join the aptamer constructs together, we utilized the NUPACK software and ensured that the linker regions were as linear as possible see Engineering Success - Diagnostic.
Our team also developed the linker regions with consideration for future steps of the diagnostic workflow, accounting for the requirements of the CRISPR-Cas12a system to include ‘TTTV’ PAM sites for the Cas12a endonuclease.
Bridge Sequence
Once we finalized the design of our aptamer constructs, we then created a 20 nucleotide ssDNA bridge sequence (BBa_25RZ6FJ4) that is complementary to ten base pairs of both left and right linker regions (Lambert iGEM, 2025c).
PDL PCR Primer Design
PCR Primer Set 1.1
To begin our experimentation, we manually designed PCR primer sequences by using Addgene’s PCR primer design recommendations (Addgene, 2019). The PCR primer design restrictions set by Addgene are:
- Length: 18-24 nucleotides
- GC content: 40-60%
- Start or end with 1-2 G/C pairs
- Tm (Melting temperature): 50-60 ºC
- Primer pairs should have a Tm within 5 ºC of each other
- Primer pairs should not be complementary to one another
Primers were cross-validated within the target region to ensure that the sequences did not have multiple binding sites. We also screened each of our primer sequences (BBa_25HV7AXA and BBa_25GA8G6C) using NUPACK to reduce the likelihood of complex secondary structures see Engineering Success - Diagnostic.
PCR Primer Sets 1.2-1.3
Because PCR primers 1.1 exhibited a limited degree of success, we used IDT’s built in PrimerQuest™ software tool to design additional PCR primer sets 1.2 (BBa_25SMX70F and BBa_258FI332) and 1.3 (BBa_25BEQMUZ and BBa_25DECTTF ). We further validated the primer sequences by ensuring that the candidates were generally consistent with Addgene’s requirements as well see Engineering Success - Diagnostic. Each forward and reverse primer set results in a different PCR amplicon, ranging between 150-250 base pairs (see Fig. 4).
PDL Experimentation
For our experimentation this year, we ran PDL in vitro using a commercial protein sample. Aside from the standard molecular biology reagents, the only specialized hardware necessary was an incubator to maintain reaction conditions during ligation see Experiments.
Experimental Modifications
Our team modified the protocol for PDL from Fredriksson et al., 2002 by introducing a T4 DNA polymerase step to convert the ssDNA ligation product into dsDNA, providing a more compatible template for the downstream Recombinase Polymerase Amplification (RPA) step (Lobato & O’Sullivan, 2018; see RPA Primer Design).
Validating PDL with PCR
Our experimentation aimed to validate the success of our PDL assay by amplifying the dsDNA product using PCR. This would allow us to isolate the PDL step and confirm that the aptamers were successfully binding to CspZ and producing an output that could be used by RPA.
To assess PDL’s specificity and reduce the possibility of false positives, we included multiple control reactions:
- No aptamers: Since a dsDNA product requires specific aptamer binding to the target protein, reactions without aptamers illustrate that PCR amplification does not occur in their absence
- No ligase: Reactions without ligase were also run as a negative control to confirm that the ligation step is necessary for formation of the product and amplification
Initial attempts were unsuccessful, as no detectable PCR product was observed (see Fig. 5).
Adjusting Protein Concentration
We began our experimentation with a high protein concentration under the assumption that more protein would improve the assay’s efficacy. We utilized a commercially available CspZ control protein from Rockland Immunochemicals, Inc. and performed a 1:100 dilution, resulting in a concentration of 10 ng/µL (Rockland Immunochemicals, Inc., 2024).
However, our computational modeling found that the blood borne CspZ concentration only reaches a maximum level of 2.00 × 106 molecules/µL (see CspZ Protein Concentration; see Engineering Success - Diagnostic). The 10 ng/µL concentration we had been using yielded 8.88 × 1010 molecules/µL, which was significantly greater than the calculated in vivo concentration.
We hypothesized that the high concentration of protein used was inhibiting the assay because the protein concentration overwhelmed the significantly lower aptamer concentration. To address this, we tested at a lower protein concentration that was more consistent with conditions reported in literature (see Fig. 6).
Despite adjusting the protein concentration, amplification of the PDL still failed, suggesting that additional factors were likely limiting product formation.
Optimizing Aptamer Concentration
After realizing that lowering the protein concentration did not allow us to visualize a successful PDL result, we used our deterministic ODE (ordinary differential equation) model to reveal another potential limitation. At the aptamer concentrations used by Fredriksson et al., 2002, our model predicted that PDL would only yield ~0.6 copies of dsDNA, limiting reproducibility (see Fig. 7; see PDL Modeling Results).
We tested a range of increased aptamer concentration inputs in our ODE, and decided upon using 20 nM of aptamers instead of the previous 20 pM, which would achieve ~800 copies of dsDNA (see Fig. 8; see PDL Modeling Results). We also began including an additional negative control lacking CspZ to show that the aptamers and ligase would not create a product in the absence of a protein sample.
This approach succeeded, as the lanes on the gel for our experimental sample showed bands at ~200 bp, which is consistent with the expected PCR product’s length of 216 base pairs with PCR primer set 1.2 (Lambert iGEM, 2025c; see Fig. 9; see PCR Primer Sets 1.2-1.3). In addition, the negative controls produced no bands, confirming the absence of unintended amplification.
These results revealed that the higher aptamer concentrations were able to successfully produce a visual PCR output, demonstrating how we had previously not been generating enough copies of the PDL product for amplification, as consistent with our ODE model for PDL.
Testing Protein Concentration Over Time
In the bloodstream, the concentration of CspZ protein changes over the course of infection as the protein naturally degrades over time. Since LANCET is intended to remain viable even at later stages of Lyme, we needed to confirm that the assay could still function at lower protein levels.
To better approximate physiological conditions, we additionally tested the expected protein concentrations at 2, 25, 50, 75, 100, and 150 days pi from our modeling curve for CspZ presence over time (see Figs. 2, 10; see CspZ Protein Concentration). All final reactions, including controls, were repeated in triplicates to ensure that results were reproducible and statistically significant.
Our results qualitatively confirmed that we were able to successfully create a PDL product and amplify it using PCR. The gel exhibited a fluorescent band at the expected ~200 base pairs, indicating a dsDNA product near that length was produced in our assay. Therefore, we realized that LANCET is functional for at least 50 days, as bands were observed in samples with protein concentration up to 50 days pi.
By using the ImageJ software to analyze mean intensity and area covered by the bands, we were also able to quantify the success of our PDL assay (see Fig. 11; see Engineering Success - Diagnostic).
Results from ImageJ analysis quantitatively confirmed that our PDL system remains functional for at least 100 days pi (see Fig. 11). The intensity of the bands was the highest at 2 days pi, immediately after the CspZ protein first becomes accessible in the bloodstream. Although the signal magnitude dropped over the 150 day period, with a loss of assay activity after 150 pi, our results demonstrate that LANCET is viable for extended periods of time.
All negative controls (no aptamers, no ligase, no protein) produced significantly lower intensities, which further validated our assay’s specificity and reduced the possibility of false positive results.
With these optimizations, our findings demonstrate that LANCET can reliably and specifically detect protein concentrations in the blood, even in low volumes, across multiple stages of Lyme, supporting its potential for integration into the downstream pipeline of our diagnostic (see Diagnostic Overview).
